PIC18F26J50 with internal oscillator and USB

[czsimon]

Hi,

I am having a problem with the PIC18F26J50, specifically with USB. I was perviously using the PIC18F14K50 with great success but have run up against a brick wall with this chip.

My hardware is pretty straight forward, I have a 1k5 running from D+ to the Vusb line and I have put a scope on both the D+ andD- and can see pulses when I plug the USB cable in, but nothing else.

Im pretty sure its something I am doing wrong with the oscillator configuration, as when I plug the USB cable in my windows host detects it but does not enumerate it, I get the hardware error message. I have an LED hooked up to an output to check if the program is actually running.
Reading other posts on the forum would indicate to me I have selected the wrong frequency or divider options.
I was hoping someone here more could take a look at my fuses and tell me where Im going wrong.

Regards

Simon


Heres my test program:

[FvM]

I'm not sure, if the clock configuration is correct, but the PIC18F26J50 doesn't specify an internal oscillator frequency
accuracy as required for USB full speed operation. The datasheet tells, it's sufficient for low speed. The typical oscillator
accuracy would however meet the USB requirement of +/- 0.25%.

[Ttelmah]

USB requires _total_ clock errors, not greater the 0.25% for full speed, and 1.5%, for low speed. This error, _includes_ the error from the host, which can be up to 0.05%. So the allowable maximum error for the slave devices, is 1.44% for low speed, and just over 0.24% for high speed devices.
On 'paper', the internal clock does meet the Low speed requirements.
The USB module data sheet, specifically says:

"The 8 MHz INTOSC included in all PIC18F46J50 Family
devices is extremely accurate. When the 8 MHz
INTOSC is used with the 96 MHz PLL, it may be used
to derive the USB module clock. The high accuracy of
the INTOSC will allow the application to meet
low-speed USB signal rate specifications."

The quoted accuracy (Table 29.12), is inside what is needed for low speed operation.

The 'typical' figure could potentially run high speed, over a limited temperature range, but It is unlikely to be reliable (how 'typical' is 'typical....).

So, for low speed, go ahead, but if you want high speed operation, use an external oscillator.

Best Wishes

[czsimon]

OK, I will try the slower clock for low-speed USB. My application isnt speed sensitive so it wont be a problem if its all I can use.

Thanks

[FvM]

To make the device enumerate as low speed, you also have to change the resistor configuration.

[czsimon]

I managed to figure out the problem! I had a bad connection on the Vusb pin.
It now works fine in both full speed and low speed from the internal oscillator. Thanks for everyones help.

[Ttelmah]

Don't rely on the full speed operation though....
The 'point' is that the data sheet says the chips will 'typically' meet the timings needed for full speed, but they only _guarantee_ timings to be good enough for low speed. If you go full speed, you ar pretty certain to find devices going wrong, if the temperature gets a bit hot/cold, or with individual chip batches. If this is a commercial product, consider the problems that this is likely to cause, and whether it is worth taking the risk....

Best Wishes

[gabogt3]